May Editorial      

A new technique to study the age of ancient Antarctic ice.
(An article about Krypton that never once mentions Superman)

Knowing the precise age of polar ice would give scientists insights into climate changes during earlier eras in Earth's history, and in turn that data could help to build models of future climate change. But that first step, accurately dating ice, is not easy - and it gets even harder if the ice is ancient. Ancient ice is usually brought up for study in the form of drill cores. Shallow cores, or the upper part of deeper cores can be dated exactly by counting individual layers. Each layer represents a year's snowfall compressed into ice. The upper layers are well differentiated because of changes in the chemical composition of the ice or in the nature of the ice. The differences may also be isotopic. For example, snow from colder periods has fewer heavy isotopes of hydrogen and oxygen. However as a core is drilled deeper into the ice, the layers start to thin out. Eventually this reaches the point where individual years cannot be distinguished. So to date the older ice more precise techniques are needed. Over the years scientists have tried a number of different techniques, including high-resolution scans of electrical resistance and models of accumulation rate variation and ice flow. But when applied to the same ice, these different techniques came up with huge differences in their estimates of the age of the ice, and these discrepancies only became greater with the depth from which the samples were taken.

To date ancient ice more accurately, methods such as radiometric dating are being investigated. The best known radiometric dating technique is carbon dating - a process that measures the decay of radioactive carbon-14, which has a constant and well-recorded half-life. The decay of C-14 is compared to a stable C-12 isotope and the ratio of those two isotopes can give us the age of the sample being measured. But carbon dating can’t be used for dating ice because carbon-14 is produced in the ice itself by cosmic rays; and anyway carbon dating can only give dates back to some 50,000 years ago. A lot of Antarctic ice is believed to be much older than that.

However, a team of scientists from Oregon State University (OSU) have recently had success with radiometric krypton dating. This process was able to determine that a sample of Antarctic ice was 120,000 years old.

Krypton is a 'noble' gas which comes in two forms - radioactive krypton-81 that has a half-life of around 230 thousand years, and krypton-83 which is stable. The gas is produced by cosmic radiation hitting Earth and small amounts are trapped in air bubbles in the Antarctic ice. By comparing the ratio of krypton-81 to krypton-83 scientists can determine the age of the air in the bubbles, and thus the date that the ice was formed.(ref).

Although scientists have been interested in using krypton dating for some time now, the drawback is that krypton-81 exists only in small quantities. This means that a sample of ice needs to be large for sufficient krypton-81 to be examined. Therefore, it was originally difficult to date ice by this technique. However, since 2011 a breakthrough in detector technology has enabled scientists to count the number of krypton-81 atoms with greater accuracy. This has made age estimation possible for relatively small samples of ancient ice. A new atom counter called Atom Trap Trace Analysis ( ATTA) has been developed by Zheng-Tian Lu at Argonne National Laboratory near Chicago.

To give an idea of the scale of the samples needed, for a recent study researchers put several 300-kilogram (about 660lb) chunks of ice into a container and melted them to release the air from the bubbles. This air was then stored in flasks. Once the air had been isolated from the ice, the krypton was isolated from the air (at the University of Bern, Switzerland), and then sent to Argonne for krypton-81 counting.

"The atom trap is so sensitive that it can capture and count individual atoms," said Buizert, who is in OSU's College of Earth, Ocean, and Atmospheric Sciences. "The only problem is that there isn't a lot of krypton in the air, and thus there isn't much in the ice, either. That's why we need such large samples to melt down."

The present study has already shown krypton dating to be a precise and practical method of determining the age of ancient ice. The group at Argonne is continually improving the ATTA detector, and hope that an ice sample as small as 20 kilograms may be sufficient in the near future. This would make krypton dating a very viable tool in the study of ancient ice. Indeed scientists believe that this technique will allow them to successfully date ice as old as 1.5 million years. Christo Buizert, the lead author of the latest publication, explains that this is important because in what is known as ’the Middle Pleistocene transition’ there was a shift in the frequency of ice ages. During the past 800,000 years the Earth is thought to have shifted in and out of ice ages every 100,000 years or so, but there is evidence that before this, shifts happened every 40,000 years..

"Why was there a transition from a 40,000-year cycle to a 100,000-year cycle?" Buizert asks. "Some people believe a change in the level of atmospheric carbon dioxide may have played a role. That is one reason we are so anxious to find ice that will take us back further in time so we can further extend data on past carbon dioxide levels and test this hypothesis."

Journal Reference:
Christo Buizert, Daniel Baggenstos, Wei Jiang, Roland Purtschert, Vasilii V. Petrenko, Zheng-Tian Lu, Peter Müller, Tanner Kuhl, James Lee, Jeffrey P. Severinghaus, and Edward J. Brook. Radiometric 81Kr dating identifies 120,000-year-old ice at Taylor Glacier, Antarctica. PNAS, (2014) April 21. http://www.pnas.org/content/111/19/6876